Novago software

correlation engine

How can you transform your company's ideas into a gold mine ? How can you turn your company into an innovative company ?

The history of the numerous inventions that make up our daily life (washing machine, car, walkman, post-it notes, etc.) shows that innovation is the result of intelligent combinations of ideas, existing technologies, issues raised by customers, etc. More often than not, these combinations in part are due to chance opportunity (a meeting at a trade fair, an acquaintance, through reading, etc.).

The correlation engine, which is integrated in all NovaGO modules, encourages these combinations and speeds up the union of ideas and people that contribute to the emergence of innovations.

It is the result of seven years’ work by I-Nova, in collaboration with the CRI computer research centre of the Ecole des Mines in Paris. It has received labels from the European Commission, ANVAR, the French ministry for research, and a patent has been filed.

The 5 key functions of the correlation engine in NovaGO modules :

• Similar ideas searches, enabling high-potential projects to be built up thanks to the association of similar ideas.
• Creative search in the innovations database, to identify innovative solutions already proposed and staff, through a method combining natural language, tags and contextual proximity.
• Recommendations, making it possible to calculate an interest profile and submit relevant ideas to the user.
• Creative search in internal/external bases, other than the innovations database (customer feedback, blogs, patents, etc.).
• Clustering and mapping of information: tools that help the user to analyse a large quantity of data by creating sub-sets by contextual and semantic proximity.

 

What is the technology used ?

The correlation technology has been developed to meet one of the challenges raised by innovation management: exploiting the wealth of issues and ideas put forward by customers and by the entire company, which are today stored in customer service databases, CRM, blogs or the internal suggestion system.

All of this information has the same storage format: just a few lines of description (4 to 20 lines) and context information (author, date of publication, machine concerned, type of problem, etc.) But "semantic" and/or “statistic” data processing software solutions are only efficient if the information has high textual content.

Correlation technology was thus developed to analyse such information. It uses a “topological” algorithm that calculates distances between information. The calculation combines analysis of the textual content with analysis of the contextual information to quickly come up with similarities between information.